Back to Marketplace
guard-scanner
Static and runtime security scanner for AI agent skills, detecting 22 threat categories and blocking dangerous tool calls in real-time.
3,939downloads5installs4stars
v4.0.2
cmdopSecuritycompliance, runtime_monitoring, security, threat_detection3/2/2026
Overview
The guard-scanner skill is a comprehensive security solution for AI agent skills, providing both static and runtime scanning capabilities. It detects 22 threat categories, including prompt injection, credential theft, and sandbox violations, and blocks dangerous tool calls in real-time.
Key Features
- Static scanning: detects 135 patterns across 22 categories
- Runtime Guard: blocks 26 patterns, 5 layers, 3 enforcement modes
- Zero dependencies
- 0.016ms/scan
How It Works
The guard-scanner skill uses a combination of static pattern matching, regex analysis, Shannon entropy calculation, and data flow analysis to detect threats. It also includes a Runtime Guard hook that blocks dangerous tool calls in real-time.
Use Cases
- Before installing a new skill from ClawHub or external sources
- After updating skills to check for newly introduced threats
- Periodically to audit installed skills
- In CI/CD to gate skill deployments
Reviews
No reviews yet.